No Plan Mode — The Right Way to Use AI

• Channel: Ras Mic
• Video: Watch on YouTube
• Tags: ras mic no plan mode ai

Overview

Ras Mic argues that planning is crucial for building features with AI, and challenges the ‘no plan mode’ influencer trend with a structured approach to agentic engineering.

Key Takeaways

• Planning is one of the most crucial things that you can do when it comes to building features and apps using ai. agentic engineering can only be successful for you when you the operator have planned properly. now, when you
• Such success in building features, building apps with its client work, whether it’s job work, whether it’s my own projects, whether it’s pluto, it’s been a success for me. and that’s what i’m going to show you in today’s video.
• Is this guy at? i’m going to show you a couple pictures. i somehow ended up in the start-hand office. if you’re not familiar with ryan’s start-hand, then you’re probably not going to understand. but i’m in the office, i’m
• How most people plan. they use plan mode. and they’ll prompt saying, build x for me. they’ll turn plan mode on. it plans. they don’t even read it. and they just build, right? and a classic example of this is
• Then hit enter. and what’s going to happen is the agent has this one line of context, right? it has this one instruction. and obviously, if you’re using a good model like opus4a or g555, you know, the thinking and
• Experience, right? you should have an idea as exactly as to what you want to build. and if you don’t have that idea, my methodology will help you a lot. so how do i go about planning? my way of
• I want to build. i also discuss trade-offs, right? what if we do this? what if we do that? then after a long discussion, then do we enter plan mode? and once we’ve entered plan mode, i don’t get it
• Files? some people are saying, oh, these are useless. these md files, these plan md files are for me. they are for me to remember what’s going on. because here’s the thing, i usually work in parallel. like, meaning, let’s
• I want to keep on going. what i’ll usually do is in another tab, i’ll do the planning. and actually, i said, i have two tabs open. usually, i have one tab building a feature and then i have multiple
• And under plan. i have active apps, desktop and simplified. so if i click on, let’s say, active. why don’t i, actually, i’m going to click on apps. this is a feature i’m working on. i basically want people to
• Is a plan that was generated after tons of back and forth. not only me and the age of a me and one of our engineers at our products to be, right? so these are the type of plans i
• You want and hit plan mode, you need to start having a conversation. and in that thread, sort of built the context where the agent sort of understands exactly what it is that you want to build. and instead of
• And i deserve better. we deserve something faster, something better. something that agents can use. and i found that solution. and that is the depot ci. depot ci is a ci engine that we deserve. listen, it’s used by companies
• Commands and you’re ready to go. you don’t even need to write code to get started. but i can already hear your question. but mike, how much does it cost? how much is the pricey? look at what it costs.
• Price of one junior chicken $3, you can get over 500 minutes. that’s over eight hours of compute. listen, guys, don’t get stuck with mediocre. don’t get used to mediocre. you deserve something better. make sure to check out depot

Transcript

Planning is one of the most crucial things that you can do when it comes to building features and apps using AI. Agentic engineering can only be successful for you when you the operator have planned properly. Now, when you think of planning, you automatically think of plan mode and you’ve probably seen your favorite influencer builder talk about, I don’t use plan mode. But here’s the thing, the way I do planning, which incorporates plan mode, has yielded me such success in building features, building apps with its client work, whether it’s job work, whether it’s my own projects, whether it’s Pluto, it’s been a success for me. And that’s what I’m going to show you in today’s video. I’m going to show you how I plan and after watching this video, you’re going to be a great planner, which is going to help you build great features. Let’s get into it. Now real quick, you’re probably like, where is this guy at? I’m going to show you a couple pictures. I somehow ended up in the start-hand office. If you’re not familiar with Ryan’s start-hand, then you’re probably not going to understand. But I’m in the office, I’m in the podcast studio. I don’t know how I got here. I give all glory to God. My life has just been insane the last couple of months. All glory to Jesus. Let’s get into the video. So this is how most people plan. They use plan mode. And they’ll prompt saying, build X for me. They’ll turn plan mode on. It plans. They don’t even read it. And they just build, right? And a classic example of this is I can open curse strike and open a new tab, and I can say, hey, build me a new dashboard page that’s going to contain analytics for all my usage. And then they’ll toggle plan mode, right? Plan mode, and then hit enter. And what’s going to happen is the agent has this one line of context, right? It has this one instruction. And obviously, if you’re using a good model like Opus4A or G555, you know, the thinking and all that stuff, it’s going to help you generate a decent enough model. But in my opinion, this is not enough. The model is going to be making a lot of assumptions for you. And that’s just not good product experience, right? You should have an idea as exactly as to what you want to build. And if you don’t have that idea, my methodology will help you a lot. So how do I go about planning? My way of planning is very different. I actually have multiple back-and-forths with the agent. Once I’m confident enough with what the agent wants to build, I get understand what I have in my noggin and it has an idea of what exactly I want to build. I also discuss trade-offs, right? What if we do this? What if we do that? Then after a long discussion, then do we enter plan mode? And once we’ve entered plan mode, I don’t get it to build right away. I get it to create an MD file and a plan folder that I have, then I get it to build. Now, a lot of people might be like, oh, what’s the point of these MD files? Some people are saying, oh, these are useless. These MD files, these plan MD files are for me. They are for me to remember what’s going on. Because here’s the thing, I usually work in parallel. Like, meaning, let’s say I have this, tab work and I have another tab working. Let’s say I have like two tabs working and like the agents cooking, like I can start playing among us or watch a YouTube video. But let’s say I want to keep on going. What I’ll usually do is in another tab, I’ll do the planning. And actually, I said, I have two tabs open. Usually, I have one tab building a feature and then I have multiple tabs planning, right? I have multiple tabs planning. Now, why don’t I show you a great example. I’m going to go to the files and I’m going to show you my plan folder. So I have a plan folder here and under plan. I have active apps, desktop and simplified. So if I click on, let’s say, active. Why don’t I, actually, I’m going to click on apps. This is a feature I’m working on. I basically want people to be able to set up any sort of CLI on their Pluto instance, very quickly like an app store. You just download it. And this is the plan I generated. Now, this is not your standard plan mode plan. This is a plan that was generated after tons of back and forth. Not only me and the age of a me and one of our engineers at our products to be, right? So these are the type of plans I want because a, I know what’s going on and what it’s going to build. And b, I’m giving the agent the necessary context it needs, right? So instead of having these weak plan modes where you just tell it what you want and hit plan mode, you need to start having a conversation. And in that thread, sort of built the context where the agent sort of understands exactly what it is that you want to build. And instead of just yapping, why don’t we build out a feature together? But before we do a quick word from today’s sponsor, I have a quick question for you. Name me something slower than GitHub actions. You can’t. And frankly speaking, you and I deserve better. We deserve something faster, something better. Something that agents can use. And I found that solution. And that is the Depot CI. Depot CI is a CI engine that we deserve. Listen, it’s used by companies like post-hog like a Jane Street, like big money Jane Street. Listen, not only is it much faster than your lousy GitHub actions, but it’s fully programmable. You can get your agent to set you up in just a few commands and you’re ready to go. You don’t even need to write code to get started. But I can already hear your question. But Mike, how much does it cost? How much is the pricey? Look at what it costs. It costs 0.001 per second compute. Now I didn't understand this. So let me quantify it for you simply. So it's standard, junior chicken. According to the day's price, inflation is crazy. It's 3.69 to 3.89. That means for the price of one junior chicken 3, you can get over 500 minutes. That’s over eight hours of compute. Listen, guys, don’t get stuck with mediocre. Don’t get used to mediocre. You deserve something better. Make sure to check out Depot CI. The link is in the description down below. So what you and I are going to plan is, most of you are familiar with Pluto right now. It is basically my agent take platform for businesses. There’s a bunch of stuff you can do. You have a dedicated chat tab. You can set up tasks, you set up routines. You can give your agent a specific email. There’s 1,000 plus connectors. You can have files, index, OCR, all that stuff here. Right on the dashboard, you can connect via imesage telegram slack. You can give your agent its own card. I ordered Schwarmo with this the other day. You can give it a phone number. You can connect your bank accounts and not only that, you can also talk to it. Hey Pluto. Hey. You can do stuff like that. So there’s a lot of cool stuff with Pluto. Now what I want us to plan together is a companion app. I want a mobile companion app. Now I don’t want it to be f***er one-on-one with the desktop app because it’s just impossible or the web app. So let’s build this companion app together. So I’m going to say, hey, so I want to build a companion mobile app. For Pluto, for people who use Pluto. Now I don’t expect it to have all the features and be one-to-one with the desktop or the web app. But I want it to be useful enough where people can use it, chat with it, get notifications and all those type of things. I want you to brainstorm some ideas on what a great companion app would look like and share with me your thoughts. So I do this. I don’t do it in plan mode. We are having a conversation. Now, if the model gets jumping and tries to build, I’ll tell you to shut up, I’ll tell you to stop. But right now, what I want is for me and the model to chat and to plan. In the meantime, you all hear about this, you all hear, there’s a broke my heart, man. How is JJ going to leave this side, and how am I going to watch side-men among us? And there’s no JJ doing dumb stuff. Like, I’m genuinely heartbroken. I thought it was a prank at first, but it’s not, if anything, this is actually how I feel. I’m going to play this video for your real quick. It’s funny by Harry. One thing in life will always be the people in this room. Right here, right now, salute me from there. It’s funny, now it’s hard to feel the heart. Because we’re brothers, brothers, brothers. We can do whatever we can, and we can do whatever we can. Thank you. Fall, fall, fall, fall, fall. But yeah, I’m a Miss, JJ being the side-men. All right, so our plan, our first iteration has been done. Well, it’s still streaming, but it’s basically done. Let’s start reviewing from the top. And actually, it’s asking some questions. Let’s review first from the top. So it’s saying the mobile is the agent and box are more control, not the cockpit. Yeah, that’s fair. Don’t try to shrink the dashboard. The companion app jobs to reach me. Let me green line the agent. Let me talk to it, keep me aware. Everything flows. The killer moments, your differentiators. These are the things that make the app genuinely useful to use daily, and they map directly to data models you already have. Approved from lock screen, the hero feature. You already have agent card approval requests, a mount merchant purpose expires that a push like Pluto wants you to spend $42 an open AI. Ooh, that’s fire. Approved and denied, see, OK, we’re talking. Action draft proposals. Action draft, basically, I explain this in a previous video. I think it’s called Cloud Core. My agent’s better than Cloud Core, or whatever, I’ll put it right here. Basically, when you communicate with the Pluto agent, there is a proof. There is draft, and there’s a block of proof, meaning it just does the task. Draft, meaning it asks you like a approval. Let’s say it’s sending an email or spending some money. That should be human in the loop. You should check for that stuff. The third one is actually just block the third one is deploying a sub agent. So let’s say, and analyze the task and say, OK, this is a heavy task. This will block the main agent. So let’s deploy a sub agent. And the final one is block, right? If you’re doing some weird or nonsense things, it’s not going to do a for you. So that’s great. It then says, task done, block the needs you, right? That’s great for the Canban task. Daily briefing routine deliveries. I love that. Voice first, capture, quick capture chat on the phone, push the top. I love this. So says what I put in MVP version one,

(Transcript truncated — full length available on YouTube)